The Paris AI Summit: A Diplomatic Failure or a Strategic Success?

The recent AI Action Summit in Paris might appear to have fallen short of expectations at first glance. Unlike the Bletchley (2023) and Seoul (2024) summits, which produced ambitious declarations, the Paris gathering concluded with a more modest statement that lacked the support of two key players: the USA and the UK.

Diplomatically speaking, this could be seen as a failure, as reported by mainstream media. Yet, dismissing the Paris summit as disappointing would mean overlooking its major achievements. The summit was a success that brought much-needed clarity to the global AI debate and marked an important shift in how the world approaches AI.

The new bar for AI clarity was set in most of the statements from JD Vance, US Vice-President, announcing the end of a ‘longtermist’ approach to AI safety and Von der Leyen, President of the EU Commission, outlining the EU’s cards for the forthcoming AI race.  

From AI safety to innovation and public good

France faced a challenge in steering the AI summit away from the safety-focused agenda that dominated Bletchley and Seoul. The Paris Summit signalled a deliberate reduction in ambition, reflected even in naming the final document a ‘statement’ rather than a ‘declaration’. This shift was not a step back but a leap forward. The summit reframed the conversation around AI, moving away from speculative long-term risks and toward immediate, tangible issues like innovation, jobs, and public good.

US Vice-President Vance’s speech epitomised this shift. He criticised the longtermist approach, arguing that it often serves the interests of incumbent tech giants rather than the public.

‘When a massive incumbent comes to us asking for safety regulations, we ought to ask whether that regulation is for the benefit of our people or for the benefit of the incumbent.’

This marked a turning point, clearing the fog around AI governance and redirecting attention to the here and now—how AI is impacting jobs, economies, and societies today.

The end of Bletchley’s AI risk focus

The evolution from Bletchley’s cautionary tone to Seoul’s balanced approach reached its culmination in Paris. The Paris Statement gave, in diplomatic terms, a very weak recognition of AI safety, a pillar of the process itself. It merely ‘noted’ the voluntary commitments on AI safety, a nuance of diplomatic language that falls just short of ignoring and far from endorsing, supporting, or encouraging.

This shift is evident in the language of the statement: while ‘safety‘ and ‘risk‘ dominated Bletchley and Seoul, Paris emphasised ‘innovation,‘ ‘governance,‘ and the ‘public good.‘

Table: Evolution of language of the AI Summits

Document	Total words	Security	Safety	Risk	Future	Trust	Innovation	Governance	Open
Bletchley Declaration (2023)	1,650	1	18	15	2	3	3	4	0
Seoul Declaration (2024)	1,800	2	22	9	3	5	10	10	1
Paris Statement (2025)	1,200	1	8	3	4	4	7	6	4

The end of the ‘Bletchley AI safety narrative’ will have far-reaching consequences. The AI safety industry has grown, with numerous organisations, researchers, and activists involved. AI safety now faces a moment akin to the fact-checking industry when X, particularly Meta, changed their content policies by dropping fact-checking and refocusing on user-driven community notes.

Rebalancing the AI risk topography

While ending the predominant focus on AI long-term risks, the Paris Statement reinforces short-term risks, including jobs and consumer protection. It includes an explicit and strong call for ‘avoiding market concentration’, a growing middle-term risk in our AI risks taxonomy.

May 2023

January 2024

January 2025

AI sovereignty reinforced

The Paris Statement stresses that countries must have ownership of their transition strategies. The language is clear by stressing ‘must’, as opposed to the more typically used diplomatic ‘should’. This spirit resonates throughout the document, with, for example, a call to develop AI models in compliance with countries’ frameworks.

Bottom-up AI via inclusion and open-source models

Although not explicitly mentioned, bottom-up AI is an underlying concept in the calls to ‘reinforce diversity of the AI ecosystem’ and an ‘inclusive approach’. The Paris Statement calls for open AI models. During the summit, France celebrated the success of Mistral, their flagship open-source model.

Multilateral and multistakeholder AI governance

The Paris Statement got the dynamics between multilateral and multistakeholder governance, often wrongly put into a binary framing of one being against the other. On the multilateral side, the summit reinforced the role of the United Nations by mentioning the Global Digital Compact (GDC) and major UN meetings and referencing the World Summit on the Information Society (WSIS). It was refreshing when the UN was under severe attacks and pressure.

The multistakeholder approach was mentioned three times in a rather short document. It also explicitly indicated energy and AI.

Focus on the public good, development, and the critical role of the UN

The Paris Statement calls ‘to narrow the inequalities and assist developing countries in AI capacity building’. It is backed by a public-interest AI platform focusing on AI as a digital public good. The statement calls for AI in ‘the public interest of all, for all, and by all’.

Public good came to the centre of the summit discussion. This marked a significant departure from the tech-centric narratives of the past, recognising AI as a tool that must serve societal needs rather than corporate or speculative interests.

Environment and energy

Unsurprisingly, Paris, a city associated with the climate change pact, focused on questions of sustainable environment. Climate was left out of the final statement. Even the environment was refocused on energy, choosing a less ideological and more practical approach to the interplay between environment and AI.

Focusing on ‘energy-friendly AI,’ one can see French handwriting—a country with a strong nuclear energy industry that passed through recent energy and environment transitions rather smoothly (no closing nuclear plants, low dependence on Russian gas).

AI economy and industry

The Paris Statement is bold on investment and shy on the AI industry. The summit saw announcements of a US$300 billion investment in AI, mainly showing that France and Europe are ready to compete with the USA and China.

The Paris Statement does not repeat public-private partnership, which has been typical for all AI documents. With a shift towards public good, the Paris summit was very critical of AI monopolies, consumer protection, and the misuse of intellectual property to develop AI models. One can say that AI tech giants should not be particularly pleased with the summit outcome.

This starkly contrasts Vice-President Vance’s strong pro-business statement, which called for as few obstacles as possible for AI tech giants. Here lies the major emerging crack between the USA, home to the main tech companies, and other actors, including the European Union.

AI impact on jobs and workplaces

The future of jobs featured highly in the summit discussions and final statement, which calls for creating network observatories ‘to anticipate better AI implications for workplaces, training, and education’.

One exciting development was the inclusion of workers in policy discussions on AI in the USA by Vice-President Vance:

‘For all major AI policy decisions coming from the federal government, the Trump administration will guarantee American workers a seat at the table, and we’re very proud of that.’

This shift in approach to social issues in a typically pro-business Republican party is worth mentioning and following in the coming years.

A new chapter in AI governance

The Paris AI Summit may not have produced a grand declaration, but it achieved something more important: redefining the global AI agenda. By shifting the focus from speculative risks to innovation, jobs, and public good, the summit laid the groundwork for a more pragmatic and inclusive approach to AI governance. The absence of US and UK support reminds us of the challenges ahead but does not diminish the summit’s achievements.

In the grand arc of AI governance—from Bletchley’s caution to Seoul’s balance to Paris’s forward-looking vision—the Paris Summit represents a decisive break from the past. It was not a failure of diplomacy but a triumph of clarity and purpose. As the world grapples with the transformative power of AI, Paris will be remembered as the moment when the conversation finally turned to the issues that matter most.